How does FBN ensure data quality?
01 de mai. de 2019
We use three checks to ensure that the analyses and data you see in FBN® are accurate:
Farmers First® - Our members have a strong interest in ensuring that their own data is reported correctly so they can take advantage of the full range of insights that FBN offers. Farmers can edit yields or report events (like crop damage) in FBN in a matter of minutes in our post calibration and manual data reporting tools, and can verify changes they submit through scale tickets.
Data Operations - FBN has a dedicated data operations team that manages data processing and cleaning. In addition to their checks on every raw file that comes into the network, FBN also deploys machine learning algorithms that throw out erroneous data points and results that appear abnormal based on the results we’ve observed in similar geographies and growing conditions.
Statistical Power - In statistics, the “law of large numbers” is a concept that describes why a dataset grows in accuracy as it grows in size. With a dataset of tens of millions of acre-events that continues to grow every day, the FBN network is the most statistically valid way for a farmer to understand how a range of factors like planting population, growing conditions, hybrid selection and others correlate with yield. We know that some monitors will over-report yield and some will under-report yield. FBN analytics run on large, aggregated datasets where these errors tend to cancel themselves out, leading to less noise in the data. As a matter of principle, the FBN platform is transparent about how many acres of data comprise the yield averages in Seed Finder, and we also provide 95% confidence intervals for the key metrics in Seed Finder, Analyze My Operation and FBN Intelligence.
Improving farm data collection practices and quality benefits all farmers. We are proud to be helping growers build a more data driven, connected and independent future.